ATL: TROPICAL STORM BILL (03L)

[littlevince]

[Gigsley]

URNT11 KNHC 191732
97779 17244 40204 55608 30500 13087 08088 /3094
41150
RMK AF303 0203A BILL OB 11
SWS = 49 KTS

[wxman57]

again, the EC has not handled the storm that well. It merely got lucky and got it right for totally the wrong reason. It has been its worst in the short term, but seems to have been doing much better in recent runs (no more 200NM speed issues)

The EC was the only model to not bring Bill into the NE Caribbean or drive it into Florida and/or the Gulf. It's been consistent from the start in showing a NW turn well east of the Caribbean. They were just talking about how well the EC has performed with Bill on the tstorms list.

[Gigsley]

URNT15 KNHC 191735
AF303 0203A BILL HDOB 23 20090819
172530 2022N 05538W 6970 03122 0038 +076 +076 126085 086 052 011 00
172600 2021N 05539W 6970 03121 0034 +076 +076 125087 088 054 008 00
172630 2020N 05541W 6970 03114 0031 +077 +077 123086 087 054 007 00
172700 2019N 05542W 6970 03112 0032 +075 +075 124089 089 055 007 00
172730 2017N 05543W 6970 03110 0034 +072 +072 126086 088 055 008 00
172800 2016N 05544W 6969 03107 0041 +063 +063 126084 088 056 010 00
172830 2015N 05546W 6970 03104 0023 +074 +074 129093 094 056 009 00
172900 2014N 05547W 6969 03100 0014 +077 +077 133094 096 056 006 00
172930 2013N 05548W 6971 03090 0001 +082 +080 136099 099 060 006 00
173000 2012N 05549W 6979 03081 9985 +092 +090 138101 103 060 006 00
173030 2011N 05550W 6969 03089 9982 +088 +088 136099 100 061 009 01
173100 2010N 05551W 6969 03085 9978 +091 +091 133100 102 062 008 00
173130 2009N 05553W 6970 03083 9985 +080 +080 133106 107 063 008 01
173200 2008N 05554W 6970 03082 9990 +078 +999 132101 104 062 010 01
173230 2006N 05555W 6969 03073 9984 +075 +075 125090 096 063 012 00
173300 2005N 05556W 6987 03051 9986 +074 +074 123093 101 062 020 00
173330 2004N 05558W 6970 03068 9995 +064 +064 125096 098 064 017 00
173400 2003N 05559W 6974 03061 9986 +068 +068 130098 100 066 017 00
173430 2002N 05600W 6963 03067 9963 +082 +082 137093 096 066 013 00
173500 2001N 05601W 6966 03064 9945 +095 +095 136091 093 065 006 00

[littlevince]

[Derek Ortt]

yeah, the EC got the end result right, but it produced a lot of high errors to achieve that result

[SEASON_CANCELED]

GFS had it spot on with its early runs when it was still a wave of africa. But took it a bit closer to the east coast than the cone is atm. One run took it right over NYC

[thetruesms]

I suppose it all depends on what your standard for success is. If all you care about is how closely the forecast line matches to what the actual line ends up looking like - the end result - then yeah, the EC did great. But for those who have a higher standard (those working on the model, for instance), there's got to be kind of a hollow satisfaction in taking the wrong path to get to the right destination.

[btangy]

again, the EC has not handled the storm that well. It merely got lucky and got it right for totally the wrong reason. It has been its worst in the short term, but seems to have been doing much better in recent runs (no more 200NM speed issues)

The EC was the only model to not bring Bill into the NE Caribbean or drive it into Florida and/or the Gulf. It's been consistent from the start in showing a NW turn well east of the Caribbean. They were just talking about how well the EC has performed with Bill on the tstorms list.

I think the EC has handled the mid-latitude features better in the long range than the rest of the models. It was the first to consistently show a pronounced and deep trough over the east coast, even though the strength and progression of the trough is now somewhat in doubt. It has had trouble with the short term motion of Bill, but not so much as to impact it's long range forecast -- probably fortuitously. If Bill truly turns N between 60-65W, I think the EC will have done an amazing job overall.

[wxman57]

GFS had it spot on with its early runs when it was still a wave of africa. But took it a bit closer to the east coast than the cone is atm. One run took it right over NYC

Not to mention the multiple runs into New Orleans/Mississippi for the GFS, and the numerous hits to Florida. GFS has been consistently too far west/left with the track.

[Gigsley]

URNT15 KNHC 191746
AF303 0203A BILL HDOB 24 20090819
173530 2000N 05602W 6973 03051 9937 +097 +097 135088 089 065 006 00
173600 1959N 05603W 6968 03053 9930 +100 +098 135089 090 064 007 00
173630 1957N 05605W 6973 03046 9930 +097 +097 136092 093 068 007 00
173700 1956N 05606W 6970 03042 9921 +100 +098 137092 093 068 005 00
173730 1955N 05607W 6967 03042 9914 +100 +098 136093 094 070 004 00
173800 1954N 05608W 6971 03032 9909 +099 +097 135094 095 070 007 00
173830 1953N 05609W 6967 03032 9911 +093 +093 133096 097 069 006 00
173900 1952N 05610W 6971 03023 9902 +095 +095 134097 097 070 005 00
173930 1951N 05612W 6965 03023 9990 +087 +999 131098 099 069 008 01
174000 1950N 05613W 6968 03012 9990 +083 +999 131098 099 070 014 01
174030 1949N 05614W 6965 03008 9990 +087 +999 135094 096 069 022 01
174100 1948N 05615W 6977 02996 9990 +089 +999 138098 102 070 017 01
174130 1947N 05616W 6966 02997 9871 +096 +096 138105 106 068 008 00
174200 1946N 05617W 6971 02988 9860 +100 +100 139105 106 067 007 00
174230 1945N 05618W 6970 02982 9852 +100 +100 138106 106 067 008 00
174300 1944N 05619W 6970 02975 9844 +100 +100 138105 106 069 007 00
174330 1942N 05621W 6968 02970 9835 +102 +102 136105 107 070 002 00
174400 1941N 05622W 6968 02964 9819 +108 +107 136101 101 075 002 00
174430 1940N 05623W 6969 02951 9810 +104 +104 137103 104 074 004 00
174500 1939N 05624W 6968 02938 9793 +107 +107 137105 106 076 004 00

[littlevince]

[wxman57]

I suppose it all depends on what your standard for success is. If all you care about is how closely the forecast line matches to what the actual line ends up looking like - the end result - then yeah, the EC did great. But for those who have a higher standard (those working on the model, for instance), there's got to be kind of a hollow satisfaction in taking the wrong path to get to the right destination.

We don't get any of the high-resolution EC data here - way too expensive. So all I can see is the free stuff on the ECMWF home page or on the Penn State E-wall. So it's not clear to me what you and Derek are saying about how it got the track right for the wrong reason.

[Derek Ortt]

the track was too slow for a long time when this first became a cyclone. Some of the errors for the first 72 hours were quite high (some 24 hour errors were over 100NM)

[wxman57]

New movement measurements:

6hr - 305 deg at 16.2 kts
3hr - 304 deg at 17.0 kts

[Gigsley]

URNT15 KNHC 191755
AF303 0203A BILL HDOB 25 20090819
174530 1938N 05625W 6970 02926 9775 +110 +108 137109 112 075 005 00
174600 1937N 05626W 6968 02920 9771 +104 +104 137111 113 075 009 00
174630 1936N 05627W 6965 02908 9990 +091 +999 138113 114 080 016 01
174700 1935N 05628W 6967 02890 9990 +084 +999 137115 117 083 023 01
174730 1934N 05629W 6972 02863 9990 +080 +999 135122 124 089 023 01
174800 1933N 05630W 6969 02848 9990 +084 +999 133128 130 093 031 01
174830 1932N 05631W 6964 02829 9990 +084 +999 135130 135 095 035 01
174900 1931N 05632W 6963 02813 9990 +095 +999 140123 129 097 031 01
174930 1930N 05633W 6982 02777 9990 +114 +999 147101 105 103 019 01
175000 1929N 05634W 6970 02774 9591 +129 +129 147098 098 102 003 00
175030 1928N 05635W 6967 02767 9576 +131 +131 148096 097 102 002 00
175100 1928N 05637W 6971 02746 9556 +136 +136 147084 088 087 001 03
175130 1927N 05638W 6967 02742 9530 +145 +133 144077 081 076 001 00
175200 1926N 05639W 6975 02717 9507 +155 +129 141068 070 069 000 00
175230 1924N 05641W 6966 02713 9483 +166 +121 136057 060 060 001 00
175300 1923N 05642W 6972 02693 9463 +175 +113 132046 051 055 001 00
175330 1922N 05643W 6971 02690 9451 +184 +096 126036 037 041 001 00
175400 1921N 05644W 6970 02683 9443 +188 +086 131030 032 032 001 00
175430 1919N 05646W 6973 02679 9443 +188 +094 133023 026 029 002 00
175500 1918N 05647W 6971 02686 9448 +187 +072 142013 016 022 001 03

[littlevince]

[CrazyC83]

135 kt FL winds just found (that is 121 kt at the surface). Need a dropsonde to find out more though.

[Derek Ortt]

SFMR still in the middle cat 3

as I said yesterday, 80% is the reduction we should be using for this storm given its wind profile. 80% of 135 yields about 105-110KT, which agrees well with the surface data (and why we are going with 110KT still)

[cheezyWXguy]

Is that 135kt FL wind? What does that equat to on the surface?